Dear all,
I was trying to calculate a double perovskite's phonon band structure and density of States using VASP and Phonopy. After the calculation, the result shows that there are quite a lot of negative frequency bands, which tells me that the structure is unstable.
But I performed Elastic calculation by finite difference method earlier (Using VASP) also, but at that moment, the result showed me that, the lattices have completed all the required criteria, and it is mechanically stable.
Now What should I conclude in this matter? Or do I need to change any steps while doing the phonon calculation?( It needs to be mentioned that, I have relaxed the structure with the following criteria:
Electronic Energy Convergence= 10e^-08 eV
Force Convergence = -10^-04 eV
Energy Cutoff = 500 eV )
Please help me out.
And also, thanks in advance!
Dear Shahran Ahmed Joy
The dispersion law for phonons is in many cases an even function of frequency, i.e. cubic, tetragonal, hexagonal lattices, this is from a theoretical point of view when looking at the Christofell equation.
Structure instability of a lattice arises in the sold called Peirers distortion, particularly for 1D crystals, and is due mainly to disorder when there is a transition from metal to an insulator.
However your question is purely numerical, let us wait for a specialist that encountered this kind of computational issue. Probably is related to the cut-off for convergency given to the parameters, both, electronic and phonon ones.
Meanwhile, see for example:
https://en.x-mol.com/paper/article/706618
Best Regards.
Many people predicts the stability of the materials using the Elastic constant method or Born stability conditions. Yes that is necessary but not sufficient.
For the dynamical stability of a material, we must look into the phonon dispersion.
Article Pressure dependent structural phase transition and observati...
I have addressed such issues in one my paper(given above) and you also find some relevant references in which the elastic constant method was showing that the material is stable but when experimentalists try to synthesize such materials that proves to be unstable and later it was confirmed through phonon dispersion that such materials are unstable.
Dear Shahran Ahmed Joy ,
You should admittedly add a bit of information about your system.
As a general rule, if your system is stable you might expect that 3 frequencies turn negative around the Gamma point in the phonon bands. This is a numerical issue, and usually these frequencies can be discarded (they are associated with rigid translations): usually they are relatively smaller than the other frequencies.
If you have more negative phonon frequencies, the point is asking why and where. The negative frequencies (or better imaginary) are the signal that the structure might be unstable. However, you said that other calculations have ruled out this possibility. Can you compare other quantities (away from the region of negative frequencies) and see if the two approaches provide similar results?
Can you compare your results with anything in the literature?
I hope this helps,
Roberto
I would suggest you to perform phonon calculations through the quantumESPRESSO. Many times I have observed that phonons are negative by phonopy but they comes out stable through phonon by QuantumESPRESSO.
The phonon band structure calculated by phonopy is usually at 0 K. To confirm the structure stability, you should calculate the phonon band structures at high temperatures using the TDEP or SED method. Some double perovskites are stable at high temperatures.
- Badges
- Science topic
- Similar topics
- Correction